The purpose of this study is to test the efficacy and safety of forty- eight hours of systemic hypothermia induced by surface cooling within six hours of traumatic brain injury. The primary end point of this Phase III trial will be to determine whether functional deficits as measured by the Glasgow outcome scale (GOS) are improved in the treatment group relative to a control population six months after the injury. The secondary hypotheses include a variety of neuropsychological measurements, determination of whether hypothermia reduces mortality within three months after severe traumatic brain injury and whether hypothermia reduces the incidence of post-traumatic and idiopathic seizures relative to untreated patients. The sample size will be 250 per group in this two armed trial that will be stratified on the basis of post resuscitation motor function. It is anticipated that the study should take somewhat more than three years to complete. This amended proposal was initially reviewed in October, 1993. It is well established that acute brain injury reduces neurological damage in a variety of situations and the investigators have already conducted - two Phase II trials of surface induced moderate hypothermia. These studies have shown an absolute improvement in the percentage of patients achieving good outcomes at three months without serious hypothermia related complications, although these studies were not large enough to show significant differences. These pilot studies did demonstrate the feasibility of the proposed procedures, and all of the centers are part of emergency medical service systems with rapid transport times. The study is designed to be able to detect a shift of ten percent in patients achieving a good rather than poor recovery. Traumatic brain injury is often an especially severe problem because the victims are young and previously healthy. The injury is unanticipated, and the victims may survive for extended periods because their general medical health is good. But, because of their mental impairments they are often unable to be productive members of society and frequently are a serious financial burden to their families. This is, of course, in addition to the serious emotional consequences of such devastating losses. The investigators cite a variety of animal models showing that hypothermia induced rapidly after the onset of acute brain injury reduces neurologic injury. Although the mechanisms of damage reduction by hypothermia are not yet adequately understood, the detailed pathogenesis of this process is not the major focus of this proposed investigation. The preliminary results of the Phase II trials are quite promising, and it is likely that the proposed study will be pivotal in demonstrating efficacy of a clinically practical treatment modality. If this study shows significant hypothermia induced reduction in brain damage, it is probable that major changes in the care of large numbers of trauma patients will follow.